From owner-freebsd-hackers  Mon Sep 30 04:26:39 1996
Return-Path: owner-hackers
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id DAA17240
          for hackers-outgoing; Mon, 30 Sep 1996 03:48:38 -0700 (PDT)
Received: from who.cdrom.com (who.cdrom.com [204.216.27.3])
          by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id DAA17140;
          Mon, 30 Sep 1996 03:48:30 -0700 (PDT)
Received: from covina.lightside.com (covina.lightside.com [207.67.176.1])
          by who.cdrom.com (8.7.5/8.6.11) with ESMTP id XAA19236
          ; Sun, 29 Sep 1996 23:06:54 -0700 (PDT)
Received: from localhost (jehamby@localhost) by covina.lightside.com (8.7.6/8.7.3) with SMTP id XAA15753; Sun, 29 Sep 1996 23:05:34 -0700 (PDT)
Date: Sun, 29 Sep 1996 23:05:32 -0700 (PDT)
From: Jake Hamby <jehamby@lightside.com>
To: dyson@FreeBSD.ORG
cc: Karl Denninger <karl@Mcs.Net>, chuckr@glue.umd.edu, hackers@FreeBSD.ORG
Subject: Re: PS broke again -- what has to be rebuilt to stop this?
In-Reply-To: <199609292040.PAA11764@dyson.iquest.net>
Message-ID: <Pine.AUX.3.94.960929225147.14488A-100000@covina.lightside.com>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: owner-hackers@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

On Sun, 29 Sep 1996, John S. Dyson wrote:

> > No, I well understand that -CURRENT changes.
> > 
> > That's the point of it being "current".
> > 
> > However, if I want to run a new kernel, I shouldn't have to rebuild half 
> > of the system utilities!
> > 
> I agree that it is a problem (but being honest -- it isn't HALF of
> the utilities :-)).  I am willing to work with the other architectural
> contributors for a good fix.  It is also one of my "hot" buttons.
> 
> John

Hey, why don't we build on the KernFS we already have?  Let's see:

hamby1# mount -t kernfs kern /kern	# This command always works
					# (kernfs is an LKM)
hamby1# ls /kern
bootfile	copyright	hz		pagesize 	time
boottime        hostname        loadavg         physmem         version

Between /kern/boottime, /kern/time, and /kern/loadavg, you can probably
get most of what uptime needs.  The rest of the information is not so
useful, but I think that the best solution is to take whatever kernel
structures the utilities in question need, and make the information
available in ASCII-formatted pseudo-files in /kern.  Between /kern and
/proc (which ps is already using), we _should_ have all the information
these utilities need; then they don't need to be poking around in
/dev/kmem, and furthermore, don't need to be SGID kmem! 

By the way, Linux has done this since the beginning (except that
everything is in /proc), and therefore a ps from kernel 0.99.x, in spirit
at least, will work on the latest 2.0.x kernel. 

-- Jake